The Impact of Artificial Intelligence on Personalized Learning and Student Learning Outcomes: A Critical Review

Artificial intelligence (AI) is increasingly being used to personalize learning through adaptive instructional systems, intelligent tutoring, automated feedback, learning analytics, and generative AI. These technologies offer opportunities to accommodate differences in students' prior knowledge, learning pace, and educational needs. However, technological personalization does not necessarily lead to improved learning outcomes. This critical review examines the relationship between AI-enabled personalized learning and academic achievement, knowledge retention, engagement, motivation, self-regulation, and higher-order thinking. Drawing on research on intelligent tutoring systems, adaptive learning, AI in education, and recent systematic reviews and meta-analyses, the article evaluates the potential benefits and limitations of AI-supported instruction. Existing evidence indicates that AI can improve selected learning outcomes, although effects vary according to instructional design, educational context, intervention duration, and assessment methods [1]–[7]. The review also identifies concerns involving algorithmic bias, privacy, unreliable AI-generated content, learner dependency, and unequal access. An evidence-to-design framework is proposed to connect learner information, AI adaptation mechanisms, instructional actions, and measurable outcomes. The review distinguishes direct educational evidence from transferable AI methods developed in other domains and provides an evaluation matrix for future studies. The article argues that AI should complement effective pedagogy and teacher expertise rather than replace them. Future research should prioritize longitudinal evaluation, transparent reporting, equitable implementation, and independent assessments of retention and knowledge transfer.